How We Almost Failed with Microservices
summary
A team migrated an IBM WebSphere monolith to a Spring Boot / Spring Cloud microservices setup for the hermes de web frontend, running on OpenShift with Docker/Kubernetes and circuit breakers, plus extensive logging and monitoring. The site serves about 300 million pageviews per year and creates ~50 million shipments; only 3 of ~45–55 microservices are live so far, and the main issues encountered were integration (CMS asset interfaces), heavy infrastructure-as-code (85–95%), and long, painful toolchain work that nearly caused the project to fail.
AI-written summary · about
chapters (21)
- 0:00Project intro
- 2:54Why microservices
- 5:29Architecture decisions
- 6:52Vertical services switch
- 8:38Resilience and circuit breakers
- 10:44Automated deployment
- 11:48Risk management
- 14:18System integration
- 16:47CMS problems
- 17:29Frontend integration
- 19:41FutureJS in project
- 22:13Six-month outlook
- 23:58Conflicting recommendations
- 25:03Project management struggles
- 26:04Toolchain complexity
- 27:49Logging and monitoring
- 29:54Migration benefits
- 30:17Org structure and Conway’s Law
- 31:44Frontend versioning
- 35:01Microservices count and maintenance
- 36:27Monitoring and DB interfaces
related talks
01h 01mMicroservices Gone Wrong
51mMicroservices and Microfrontends: Scaling, Learning, Growing
23mOn the road from Monolith to Micro
26mFighting the Good Fight at the Hot Gates of Micro-Services
33mAdventures with Micro Services in Rails
23mMicroservices – Diptanu Choudhury
45mDesigning, building, testing and deploying microservices
18mMicroservices Out In the Wild